Bringing balance and technical accuracy to reporting odds ratios and the results of logistic regression analyses
Jason W. Osborne
Abstract
Logistic regression and odds ratios (ORs) are powerful tools recently becoming more common in the social sciences. Yet few understand the technical challenges of correctly interpreting an odds ratio, and often it is done incorrectly in a variety of different ways. The goal of this brief note is to review the correct interpretation of the odds ratio, how to transform it into the more easily understood and intuitive relative risk (RRs) estimate, and a suggestion for dealing with odds ratios or relative risk estimates that are below 1.0 so that perceptually their magnitude is equivalent of Ors or RRs greater than 1.0. Accessed 37,451 times on https://pareonline.net from October 18, 2006 to December 31, 2019. For downloads from January 1, 2020 forward, please click on the PlumX Metrics link to the right.